Resume Job Matcher

---

Overview

Resume Job Matcher is an end-to-end NLP project that extracts skills from resumes and matches them against job descriptions to provide a skill match score. This project demonstrates hands-on experience in Natural Language Processing, MLOps, CI/CD pipelines, containerization, and cloud deployment.

The service is deployed on Railway as a live API that can process resumes in PDF and TXT formats and return matched and missing skills for any job description.

---

Key Features

• NLP-based Skill Extraction:
  Extracts technical and soft skills from resumes using a custom NER-inspired pipeline with spaCy and PhraseMatcher.

• Job Matching:
  Compares extracted resume skills against job descriptions to calculate a match score, highlight matched skills, and identify missing skills.

• Preprocessing Pipeline:
  Cleans and normalizes text from resumes and job descriptions to ensure consistent matching.

• REST API with FastAPI:
  Exposes a /match endpoint for uploading resumes and job descriptions via POST requests.

• Streamlit UI:
  Provides an interactive web interface for uploading resumes, submitting job descriptions, and viewing match results.

• Dockerized Deployment:
  Fully containerized with a Dockerfile for consistent environments and easy deployment.

• CI/CD Pipeline (GitHub Actions):

  • Automatically builds and installs dependencies on each push or pull request.
  • Runs tests to ensure reliability before deployment.
  • Demonstrates professional MLOps and automation skills.

• Supports Multiple File Formats:

  • .pdf and .txt resume files.
  • .txt job descriptions.

• Cloud Deployment:

  • Live API hosted on Railway for real-time testing and demonstration.

---

Project Structure

Resume-Job-Matcher/
│
├─ src/
│ ├─ api/ # FastAPI app
│ ├─ ingestion/ # Resume/job description loader
│ ├─ matching/ # Skill matching logic
│ ├─ ner/ # Skill extraction and normalization
│ ├─ pipeline/ # End-to-end pipeline
│ └─ preprocessing/ # Text cleaning and preprocessing
│
├─ data/ # Sample resumes and job descriptions
├─ tests/ # (Optional) Unit tests for CI
├─ tmp_uploads/ # Temporary uploaded files for API
├─ Dockerfile # Container configuration
├─ requirements.txt # Python dependencies
├─ .github/workflows/ # CI/CD pipeline
├─ README.md
└─ LICENSE

---

---

Installation

1. Clone the repository:

git clone https://github.com/theHoodguy4587/Resume-Job-Matcher.git
cd Resume-Job-Matcher

2. Install dependencies:

pip install -r requirements.txt
python -m spacy download en_core_web_sm

3. Run locally:

uvicorn src.api.main:app --reload

Visit http://127.0.0.1:8000/docs for Swagger UI.

Docker Deployment

Build and run the Docker container:

docker build -t resume-job-matcher .
docker run -p 8000:8000 resume-job-matcher

Cloud Deployment

This project is deployed on Railway: https://resume-job-matcher-production.up.railway.app/docs#/default/match_resume_job_match_post

Streamlit UI

Live app: https://thehoodguy4587-resume-job-matcher-srcapp-eiggvh.streamlit.app/

Run locally:

streamlit run src/app.py

Screenshots

CI/CD

• The project includes GitHub Actions for automated testing and deployment:

• Runs tests for skill extraction and preprocessing

• Builds Docker image on push

• Deploys automatically to Railway

Tech Stack

• Python 3.10

• FastAPI for API

• spaCy for NLP / NER

• Docker for containerization

• GitHub Actions for CI/CD

• Railway for cloud deployment

How it Works

1. Upload a resume (.pdf or .txt) and a job description.

2. Pipeline processes text: cleaning → skill extraction → normalization → matching.

3. Returns a JSON with:

• Resume skills

• Job description skills

• Matched skills

• Missing skills

• Match score

Future Improvements

• Add authentication and role-based access for the UI.

• Expand NER model for more advanced skill extraction.

• Integrate resume scoring and ranking for multiple candidates.

• Add multi-language support.

License

MIT License © 2026 Senitha Gunathilaka